Testing the temporal stability and longitudinal measurement invariance of the Patient Health Questionnaire-4 across six years with the Socioeconomic Panel-Innovation Sample

Abstract

The Patient Health Questionnaire-4 (PHQ-4) is a standard screening instrument for depression and anxiety. Based on its short length it is frequently used in large panel studies. While there is robust evidence for several aspects of the PHQ-4's reliability and validity, the knowledge about its temporal stability is very limited, with few studies only providing findings that are affected by small sample sizes, limited generalizability, and short-term and unstandardized retest-intervals. We aimed to narrow this gap in the literature by analyzing the longitudinal measurement invariance and retest correlations of the PHQ-4 by using data from the 2012 and 2018 waves from the representative German Socioeconomic Panel-Innovation Sample (N = 749). Longitudinal measurement invariance analyses supported scalar invariance of the PHQ-4's latent measurement model and we found robust test-retest associations for the manifest scores (rs between 0.31 and 0.37) and in latent variable analyses (βs between 0.51 and 0.57) for the 6-year lag. Our findings further support the reliability (in terms of expected stability) and structural validity of the PHQ-4 over time and support its use as a screening tool in longitudinal panel studies.